feat(vision): add OCR contracts (#355)

* feat(vision): add OCR contracts

* fix(vision): harden OCR contracts

* refactor(vision): use OCR terminology

Cargo.toml

@@ -52,6 +52,14 @@ env_logger = "0.11"
 # Optional native page rendering for OCR pipelines. PDFium is loaded at
 # runtime, so enabling this feature does not link or download a native library.
 firecrawl-pdfium = { version = "0.1.0", optional = true }
+# Small support crates used only by the opt-in model cache. Model files remain
+# external and are never embedded in pdf-inspector artifacts.
+dirs = { version = "6.0", optional = true }
+fs2 = { version = "0.4", optional = true }
+sha2 = { version = "0.11", optional = true }
+
+[target.'cfg(all(windows, not(target_arch = "wasm32")))'.dependencies]
+windows-sys = { version = "0.61", features = ["Win32_Storage_FileSystem"], optional = true }
 
 # Browser builds use JavaScript randomness for encrypted PDFs and embed the
 # bundled CMaps because there is no filesystem at runtime.
@@ -65,7 +73,9 @@ tempfile = "3.3"
 [features]
 default = []
 python = ["pyo3"]
-render-pdfium = ["dep:firecrawl-pdfium"]
+vision = []
+model-cache = ["vision", "dep:dirs", "dep:fs2", "dep:sha2", "dep:windows-sys"]
+render-pdfium = ["vision", "dep:firecrawl-pdfium"]
 
 [[bin]]
 name = "pdf2md"